Fire_Storm

hey so this problem started happening recently, the problem is eather my video card or my power supply, im not sure
it first started happening when i tryed to load up mirrors edge, right after loading screen, when the resolution was about to change, i assume it would turn off. stopped trying to play that game, but it did it again with an old school game jets and guns, at this point i was thinking it was a video card problem, though i could play other games fine, like tf2 full graphics for hours and it would be fine, it also turned off right after the directx7 test in dxdiag

more recently now though, i cant play games for a period of time, it turns off on in the middle of playing tf2 and cs source, and after it turned off it wouldnt turn on for a long time
i would press the power button and it would light up for a split second, and turn off, i have to hold the power button down for 5 seconds to actually turn it off before i can hit the on button again, keep trying over and over, eventually, itl start turning on for a few seconds, then turning off, when this happens i know its about to turn on, so a few more tries with it turning on for a few seconds, then eventually itl turn on and stay on it. this generally takes about 10 minutes
most recently it happened watching a full screen video on kotaku, so flash based this time it took about an hour and half before it would turn back on


so at first i was thinking it was the video card because it was only donig it during graphics intensive stuff, maybe its overheating, and wont turn on until it cool enough, though before this my video would vpu recover and be fine and not have to turn off and ive had it playing a lot harder games to run than cs, be fine previously
but this not turning on thing seems to point to power supply, especially since recently it was during watching a flash video, because most of the work when watching flash is done by the cpu not gpu.
i know its not my cpu overheating because i rendered some videos, putting my cpu at 100% for a while and it didnt turn off.


my computer specs are
AMD Athlon64 X2 4400+ Dual Core Processor
2x 1024MB OCZ DDR PC-4000 Dual Channel Gold Edition
the power supply is the one that comes with
Antec Sonata II Mid Tower with 450Watt SmartPower Power Supply
video card is radeon x1900xt
any help would be appreciated
oh ya heres a screenshot of sensor program at basically idle
http://img211.imageshack.us/img211/7099/sensors.png
my +3.3voltage is low? is that normal?

edit: also id rather not do some diagnostic stuff where i force it to turn off because im afraid it wont turn back on next time, and exams are coming up

Tumbleweed36

The power supply could very well have some difficulties, but I need to cover something I passed over when I got called to dinner in the middle of reading your post and therefore, missed the sensorview stuff. When I came back, I overlooked it.

You are way too hot if your sensors are correct. While what you tell as symptoms could very well be a bad power supply (shutting down, not starting), you need to get the thermal paste changed in this rig before you do anything else.

Fire_Storm

thanks for the info, so would the shutoffs after the loading screens, be caused by the power supply because thats when the resolution changes, and requires an extra bit of power to the video card which the power supply couldnt handle?
and during games because extra fans have to spin up to cool?
because other than that it will stay on for days

also the only other hardware i have in there is a dvd drive, and 3 sata hard drives, and its been fine up until this point, would i really need a 650W power supply?

also is it bad that my +3.3 voltage fluxuates, and gets pretty low, or is that normal

Fire_Storm

really? the fan on this case also cut out a while ago, meaning to get it changed i just have the side open,
what are more normal numbers?

Tumbleweed36

I wanted to go back and look at the Sensorview map again to answer some of your questions, but can't get it to load for some reason right now. I will try again in a few and report back to you when I get it loaded. Hope you got my explanation PM.

Tumbleweed36

Finally got it to load....Your idle temps should by no more than 45c (and that is pretty hot) and your under stress temps should be not more than about 63c.

The 3.3 rail indicates that you most likely do have a power supply issue. Here is the skinny.....if the power supply is the problem, it could be causing heat, so you need to look in that direction and get something like a 650 watt Corsair supply to put in there. You do need a 650 Corsair in there IMHO to make sure that baby is getting enough power.

http://www.newegg.com/Product/Produc...82E16817139005


Also, with those temps, I would change that thermal paste for Artic Silver.

Just to be safe.....check your temps in the bios to make sure the Sensorview is not giving you bad readings. Some times, those sensors can be faulty (quite common),
While in the bios, also check the voltages and report both back here and we can look closer at the most pressing issue.

Fire_Storm

in bios cpu was 61 motherboard was 39

vcore votange was 1.375
3.3 was 3.28
5v was 4.977
12v was 12.10

i guess i should mention about my video card, i noticed my video card was running rather hot when playing games, up to 100 degrees, at that point it would usually vpu recover, that was on intense games like assassins creed, usually around 90 degrees for cs source or tf2 i think, and althought it was hot, it ran fine and i never had any problems and idle it would be about 65-70 degrees

http://ixbtlabs.com/articles3/cpu/vp...a-topmount.jpg
i did get a new cooler for it though, the hr03, which is pictured above now idle is about 45 and doesnt go above 75 playing a game, thats the gpu, however the stock cooler had the fan cooling the video memory, the hr03 , just has small heatsinks on each memory chip, and i suspect does a much worse job cooling, as i would vpu recover on some games, even though the gpu temperature low. and i cant do anything about it now, because the heatsinks are thermal expoxied on
edit: though things were running fine with this setup for months before this problem started occuring

Tumbleweed36

Your voltages are most likely more correct with the bios as are your temps. Those voltages look fine at least, even if you appear to be underpowered.

I still think you need to change that thermal paste ASAP, because those temps are way to hot for the CPU. Be sure you clean off the old before you put on the new and pay special attention to how much paste they tell you to put on there:

Applying Artic Silver Thermal Paste

That might stop the shut downs and then we would know more about the health of the power supply, although I am still very certain you are underpowered by quite a bit.

If you keep running the rig at those temperatures, it is possible you might harm something, so we need to get those temps down.

Fire_Storm

thanks, i will likely do that next week, as this week i have exams and need my computer to be on, i have an antec true480 power supply that i think im going to try in there as well

Fire_Storm

k so ive confirmed its the power supply, my pc now wont turn on anymore unless i unplug one of my hard drives. obviously this one is faulty
but i ran my rig through antecs power supply calculator, and according to its calculations i only need about a 360 watt power supply for my rig, now i realize that tool probly isnt the best so even accounting for a large magin of error, a 450-500 watt power supply should be enough shouldnt it?
and before this one gave out on me it was just fine this one just happened to be low quality i guess

Tumbleweed36

With that video card I recommend a 550 or 650 watt power supply. Here is a great deal on a top of the line 650 watt power supply and it is only about $5 more than a 550 watt supply from the same company:

http://www.newegg.com/Product/Produc...82E16817139005